Researchers at Carnegie Mellon University have developed a low-cost way of detecting people through walls by analysing two WiFi routers that depict a human’s 3D shape and movements using artificial intelligence (AI).
Observation Through Walls
A deep neural network called DensePose maps the phase and amplitude of WiFi signals to UV coordinates. According to researchers in a recent study, this is what occurs when the surface of a 3D model is projected onto a 2D image for mapping a computer-generated image.
Scientists from Imperial College London, Facebook AI, and University College London developed the algorithm. Additionally, scholars from Carnegie Mellon University created the actual tool
Researchers at Carnegie Mellon University have created a system that accurately maps the postures of many individuals, according to a report by Vice and ZDNet. Instead of more expensive RGB cameras, LiDAR, or radars, it just employs a low-cost, widely accessible 1D sensor, such as WiFi antennae.
They were able to successfully localise an object in space and use WiFi to identify people and their body positions.
According to the article DensePose From WiFi by researchers Jiaqi Geng, Dong Huang, and Fernando De la Torre, “the results of the study reveal that our model can estimate the dense pose of multiple subjects, with comparable performance to image-based approaches, by utilising WiFi signals as the only input.”
In a Home Setting, The researchers contend that their WiFi approach to identify people in houses may be valuable in in-home healthcare, where patients may not want to be watched with a camera in places like the bathroom or other sensors and monitoring devices.
The majority of houses in developed countries now have WiFi, according to the study. They assert that this technology may be scaled to keep tabs on the well-being of senior citizens or to spot shady activity in the household.
The WiFi monitoring system is notable in that it is unaffected by low illumination or physical obstructions like walls.
The team claims that their device is a low-cost and “privacy-friendly” substitute for cameras and Lidars despite the fact that it does not provide a clear image of the object, according to The Independent.
With just standard components accessible in most homes, the setup is quick and affordable. It is only an extra benefit.
The two TP-Link versions used in the study were just around $30 each, but the most costly WiFi routers may cost up to $700.
“We think that in certain situations RGB visuals can be universally replaced by WiFi signals for human sense. Wi-Fi-based interior monitoring systems are mostly unaffected by illumination and obstruction. Additionally, they safeguard people’s privacy and the necessary equipment is reasonably priced, “scientists said.